June, 1920] Origin of Cerebral Ganglia 309* 



reasoning could be applied to the dorso-lateral placodes, and 

 yet, to m^'' knowledge, this has not been done. It becomes still 

 more unsatisfactory if applied to the neural crest. The idea is- 

 attractive, however, but certainly receives no confirmation from 

 evidence furnished by either the structure or the mode of 

 development of the epibranchial placodes. They are simply 

 ganglion forming structures, as are the dorso-lateral placodes, 

 and the neural crest. In all cases the specific sense organs- 

 arise entirely separate in position, and usually later, and very 

 frequently without any close connection with epibranchial or 

 dorso-lateral placodes. The attractive conception should, in 

 my opinion, be entirely disregarded for the more accurate idea 

 that these are simply ganglion forming structures. 



The evidence that epibranchial placodes form gustatory 

 ganglia rests upon the facts, (1) that in some types a ganglion 

 such as the visceral portion of the IXth is almost, if not entirely,, 

 placodal in origin, and in this case the visceral IXth nerve is. 

 almost exclusively gustatory in function. (2) In other types,, 

 every step in the contribution of placodal cells to the ganglion 

 can be clearly traced. (3) Every nerve, in all forms studied 

 that possess gustatory fibers, has epibranchial placodes pro- 

 portionate in size to the gustatory component. (4) No nerve 

 not having gustatory components has an epibranchial placode.. 

 The ganglia showing these characteristics are the genticulate, 

 nodosal and petrosal in all forms. The Vth ganglion and nerve 

 gives rise to no gustatory fibers, and never has an epibranchial 

 placode. 



To summarize these generalizations, it seems to me fairly 

 safe to make the following statements: 



1. That all spinal ganglia arise from neural crest cells 

 detached from the re-entering angle between the ectoderm and 

 the neural canal. These cells, of course, give rise to general 

 somatic and general visceral fibers. 



2. In the head region the same general condition exists,, 

 namely, that the neural crest and its homologue give rise to all 

 general somatic and general visceral ganglia, but the neural crest 

 in the head shows three rather distinct modes of behavior: 

 First, that which is identical with the behavior in the spinal 

 cord region; second, that in which the neural crest remains in 

 the ectoderm, and later becomes detached from this position;: 

 third, that in which the neural crest is incorporated in the neural 



